Discrimination of materials has become a standard requirement for security inspection systems. An inspection system for cargo and containers screening typically employs an electron accelerator capable of interlaced dual energy operation, and differential transmission of X-rays characterized by distinct energy spectra can be used to distinguish among intervening materials of different atomic numbers. The term “interlaced energies,” as employed herein, denotes the use of a stream of X-ray pulses wherein successive pulses are characterized by distinct energy spectra. The use of interlaced energy inspection for material discrimination is well known, with processing techniques surveyed, for example, by Ogorodnikov et al., Processing of interlaced images in 4-10 MeV dual energy customs system for material recognition, Phys. Rev. Special Topics—Accelerators and Beams, vol. 5, 104701 (2002), and in references cited therein, all of which are incorporated herein by reference. A Bremsstrahlung spectrum is characterized by its endpoint energy, defined by the energy of electrons impinging upon an X-ray target in order to generate the X-rays. Attenuation by an inspected object of transmitted radiation for two (or, on some occasions, more) distinct energies provides the input data for identifying the type of material being inspected.
Various techniques are known for generating X-rays of interlaced energies based on electron accelerators, such as described, for example, in U.S. Pat. No. 7,646,851, entitled “Device and Method for generating X-Rays Having Different Energy Levels and Material Discrimination System,” and U.S. Pat. No. 8,604,723, entitled “Interlaced multi-energy radiation sources,” both of which are incorporated herein by reference. The technologies of interlaced energy irradiation merit no further discussed here, since they are irrelevant to the present invention, other than to highlight their deficiencies for purposes of cargo inspection.
Several limitations accompany material discrimination functionality when a source of interlaced X-ray energy is employed. Using two pulses separated in time for producing one inspection data point effectively reduces inspection speed. Moreover, while the basic assumption of dual-energy techniques is that the same region of the cargo is being probed by both energies, it must be borne in mind that the cargo and the probe are typically in relative motion. Interlaced energy approaches are thus only tenable for slow moving cargo.
X-ray security inspection systems for the inspection of cargo and shipping containers typically use transmission radiographic techniques. FIG. 1 depicts a cargo inspection system employing such a technique. A fan-shaped beam 12 of penetrating radiation, emitted by a source 14, is detected by elements 18 of a detector array 16 distal to a target object, here truck 10, is used to produce images of the target object. Detector elements 18 produce corresponding detector signals processed by processor 19 to provide information relative to the material composition of the cargo and images of its spatial distribution. The thickness of material to be penetrated by the X-rays may exceed 300 mm of steel equivalent in some cases. To insure the required penetration, inspection systems typically use X-rays with a maximum energy of several MeV, currently up to about 9 MeV. X-rays in excess of 1 MeV are frequently referred to as hard X-rays or high-energy X-rays. While the invention described herein pertains to any penetrating radiation, it may be described, purely as a matter of heuristic convenience, in terms of high-energy X-rays.
Information (such as mass absorption coefficient, effective atomic number Zeff, electron density, or the spatial distribution of any of the foregoing, etc.) with respect to the material composition of the contents of objects may be obtained on the basis of the interaction of X-rays with the material, and, more particularly, by illuminating the material with X-ray beams having energy spectra with more than one distinct energy endpoint (peak energy), or by employing energy discriminating detectors. Dual energy methods of material discrimination are widely used in X-ray inspection systems for security control of hand luggage in customs and other security checkpoints.
Dual (and, more generally, multiple-) energy methods have been extended to high-energy inspection systems for cargo containers, where material discrimination is less effective due to the weaker Z-dependence of the dominant interaction.
In the practice of dual-energy inspection, X-ray transmission data of an inspected object are obtained for both energies, and processed by computer, whereupon a resulting image is displayed on a monitor, typically in a special color palette that facilitates visual identification of contraband or hazardous materials. More particularly, special computer software may identify various materials and artificial colors may be assigned to various values of Zeff.
A typical energy range for the inspection of smaller objects is below 0.5 MeV, taking advantage of the strong Z-dependence of the X-ray attenuation coefficient due to the prevalence of the photoelectric interaction (characterized by a cross-section, ˜Z4-Z5) at lower energies. In the range of 1-10 MeV, however, X-ray interaction is dominated by the Compton effect with its weak dependence of attenuation coefficient (mass absorption) on the atomic number: μc˜Z/A (which is approximately constant and equal to 0.5), where Z denotes atomic number, and A denotes atomic mass, which is to say that the mass absorption coefficient is largely Z-insensitive in the energy regime dominated by Compton scatter.
A preferred method for material discrimination entails variation of the pulse energy during the course of each single pulse, as described in detail in U.S. Pat. No. 8,457,274 (“Arodzero '274”, issued Jun. 4, 2013), which is incorporated herein by reference.
Leó Szilárd conceived of the linear accelerator (linac) in 1928, while a professor at the University of Berlin. A linac was also constructed independently by Rolf Widerøe, then an engineering graduate-student under Walter Rogowski at Aachen, at about the same time. Electrons accelerated by a linear accelerator were first used to generate X-rays at Stanford in the mid-1950's.
Some prior art methods for varying the emitted energy during the course of a pulse have required that the x-ray flux track the end-point energy. The Arodzero '274 patent, for example, states that “Concurrently with the sweeping of the endpoint energy, the X-ray flux may increase from a minimum to a maximum.” (Arodzero '274, col. 6, lines 47-48.)
US Published Patent Application 2014/0270086 (to Krasnykh), incorporated herein by reference, describes an intra-pulse multi-energy method that uses a traveling wave accelerator structure. It suggests the use of feedback to the electron gun grid voltage to compensate for X-ray flux variation during the course of a pulse. Krasnykh et al., Concept of RF Linac for Intra-Pulse Multi-Energy Scan, SLAC Pub-15943, (Apr. 18, 2014) provides further description, and is also incorporated herein by reference. The prior art mode of operation, however, could not accommodate separate tailoring of the flux and end-point energy of an X-ray pulse, even though such operation would be highly advantageous in a cargo inspection context, for example.
One of the limiting factors of inspection speed is RF-power available for accelerating. The maximum pulse repetition frequency (PRF) that a linac-based X-ray source can provide is limited by the RF source. The RF source (typically, a magnetron or a klystron) has limitations on maximum average Pav,max and pulsed Pp,max power. These two parameters define the maximum duty factor dmax, which also can be expressed in terms of PRF (f) and pulse duration tp:
                              d          max                =                                            P                              av                ,                max                                                    P                              p                ,                max                                              =                      f            ·                                          t                p                            .                                                          (        1        )            
For example, where a single energy (SE) (non-interlaced) accelerator, characterized by Pp,max, is chosen to produce the high energy (HE) pulse, with tp≈3.3 μs and dmax≈0.001, the maximum PRF would be limited to fH≈300 Hz (pps).
For a dual-energy interlaced linac, the maximum available frequency can be estimated from the equation
                                                        f              DE                        ≈                                          P                                  av                  ,                  max                                                                              P                  H                                ·                                  t                  p                                ·                                  (                                      1                    +                                                                  P                        L                                            ⁢                                              /                                            ⁢                                              P                        H                                                                              )                                                              =                                    f              H                                      (                              1                +                                                      P                    L                                    ⁢                                      /                                    ⁢                                      P                    H                                                              )                                      ,                            (        2        )            where PH and PL represent the RF power necessary to produce high (HE) and low (LE) energy pulses, respectively. If the assumption is made that tp remains the same for both energies, and that PH=Pp,max, then, for PL=PH (RF-power remaining constant for both pulses, achieved, for example, by RF-switches/regulators, manipulation of beam loading, and phase-shifting of the accelerating field),
      f    DE    =            1      2        ·                  f        H            .      That is to say, that a dual energy repetition rate of, at best, half that of the single-energy rate, may be achieved. On the other hand, if the low-energy pulses produce only half the power of the high-energy pulses,
            P      L        =                  1        2            ⁢              P        H              ,(as might be implemented using RF-generator power supply modulation, for example), then
            f      DE        =                  2        3            ·              f        H              ,which is to say that ⅔ of the single-energy pulse rate may be achieved on an interlaced energy basis.
In prior practice, both the RF-power and the injected beam were turned on at the same time (tb=0). The result of such prior art practice is shown by the dotted curve 30 plotted in FIG. 3, which represents the beam energy W vs. time t for a 6-MeV accelerating structure designed for security applications. The filling time, which is the time it takes for the electric field in the accelerator structure to decay to e−1 of its initial value, is tf,95%≈1 μs.
A well-known technical solution for reducing the filling time was described by Roger Miller, Comparison of Standing-Wave and Travelling-Wave Structures, SLAC Linear Accelerator Conference, SLAC-PUB—3935, pp. 216-21 (1986) (hereinafter, “Miller (1986)”, which is incorporated herein by reference. The Miller solution allows for creating the beam pulse with constant energy over the pulse duration. The accelerating beam turns on with delay tb that is defined as:
                              t          b                =                  τ          ·                                    ln              ⁡                              (                                                                            4                      ⁢                      β                      ⁢                                                                                          ⁢                      rLP                                                        IrL                                )                                      .                                              (        3        )            β is the coupling coefficient between an RF power feed waveguide 222 (shown in FIG. 2) and an accelerating structure 22 (shown in FIG. 2, also referred to herein as an “accelerating cavity structure”), r is the shunt impedance of the accelerating structure 22, L is the length of the accelerating structure 22, and P is the power dissipated in the accelerating structure 22, and τ is the decay time constant of the accelerating structure 22. (Thus, both numerator and denominator of the logarithmic argument have units of voltage.) Rigorously, β is defined as the ratio of power lost outside the accelerating cavity structure 22 (i.e., in the feed waveguide 222) to the power dissipated inside the accelerating cavity structure 22. If β=β0 has been adjusted so that there is no RF-power reflection from the accelerating structure 22 when the beam 220 is on, the above equation can be cast as:
                              t          b                =                  τ          ·                      ln            ⁡                          (                                                2                  ·                                      β                    0                                                                                        β                    0                                    -                  1                                            )                                                          (        4        )            where β0 is the optimum coupling coefficient,
                                          β            0                    =                                    (                                                                    I                    2                                    ⁢                                                            rL                      P                                                                      +                                                      1                    +                                                                                            I                          2                                                4                                            ⁢                                              rL                        P                                                                                                        )                        2                          ,                            (        5        )            and τ, as above, is the decay time constant of the accelerating structure 22.
As known to persons of ordinary skill in the art, the coupling coefficient of the accelerating structure 220 (also referred to as an “accelerating resonator,” or a “resonator,” or “RF accelerating structure”) to the external circuit (feeding waveguide 222) depends on the current accelerated in (and interacting with) the resonator 220. Typically, the presence of current decreases the coupling coefficient that is measured through VSWR (voltage standing wave ratio), and the phase of the reflected signal from the resonator 220. Initially (without current), the resonator 22 needs to be over-coupled and to have a coupling coefficient of greater than β=1. The optimum coupling coefficient β0 is a value that allows the resonator 22 to be matched with an external waveguide 222 at the accelerating current I. When the coupling coefficient β is equal to β0, the coupling is referred to herein as “optimal.” An exact calculation of the optimum β0 may be found by reference to Sobenin et al., Electrodynamic Characteristics of Accelerating Cavities (Eng. trans.), CRC Press, particularly at p. 121 (Eqn. 4.49), (1999), Collin, Foundations for Microwave Engineering, McGraw-Hill, (1st ed., 1992), and Gao, Analytical formula for the coupling coefficient β of a cavity waveguide coupling system, Physics Research A, vol. 309, pp. 5-10 (1991), all of which are incorporated herein by reference.